1S6 ALASKAN GLACIER STUDIES
This south distributary was traversed eastward to the very end of the nunatak to make sure of the contact with the tidal distributary, which changed so much between 1909 and 1910. The snow line was very low at the time of this journey, June 17,1910, reaching practically to the end of the land arm. Nevertheless long strips of glacier surface had been bared by the wind, and the snow was nowhere very deep, so it was clearly evident throughout the journey that there had been no change in this arm of the glacier since 1905 and 1909, except lowering by ablation. The hard coarsely-crystalline ice was not at all crevassed.
From the extreme end of the nunatak, which we had not visited before, it was plain that the area of crevassing that extends southeastward from the nunatak is of a somewhat different nature than more distant views had led us to infer. It is not an elongated dome of crevassing which we have thought was wholly due to the exposure of the continuation of the rock ridge of the nunatak. Instead it is a crevassed northeast edge of the more stagnant arm of the glacier and in 1910 stood several hundred feet higher than the crevassed surface of the tidal arm which flows to the fiord. The crevasses are in gigantic steps descending northeastward and breaking the edge of the stagnant arm. Evidently during the two and a half mile retreat of the tidal distributary, between 1891 and 1909, its surface has been lowered faster by ablation and flowage combined than the stagnant arm, where ablation works alone and where the retreat was only a half mile or so. The north edge of the stagnant portion is, therefore, breaking because of removal of support and is perhaps preparing to flow later into the tidal distributary instead of the smaller arm south of the nunatak, down which it discharged in former times. The rock ledge continuing the nunatak is doubtless here as we have hitherto inferred, but the lower ice on the north side seems primarily responsible for permitting the crevassing.
The Sea Tongue, or Tidal Distributary. The tidal distributary of Nunatak Glacier, which is far larger than the land tongue, though not now extending so far west, is greatly crevassed from side to side. When visited by us in 1905 and 1906 the sea face was less than a mile, wide, ending in a deep fiord bordered by precipitous mountain walls against which rested narrow lateral moraines which projected slightly farther than the ice front, especially in the case of the narrower south moraine. There were three medial moraines and almost no other debris on the glacier. The ice front, which rose precipitously for over 200 feet, projected farthest in the middle, and discharged icebergs almost constantly, the iceberg waves giving rise to violent surf which was doing important wave-erosion work on several miles of adjacent coast. The icebergs, mostly free from de'bris and including masses that rose 20 to 30 feet out of the water, were less numerous than those from Hubbard and Turner Glaciers.
Two streams emerged from the glacier below sea level, each about 200 yards from the glacier margin, giving rise to swirling currents of muddy water which kept floating ice away from the sea cliff, except just after an iceberg falL The northern stream, which was the larger, showed its influence ,for over a mile. Between the streams there was a mass of floating ice close to the glacier front. These streams were thought to be accumulating a large deposit of seo!iment beneath the waters of the fiord.
Recession of the Tidal Distributary. As in the case of the land tongue of Nunatak Glacier, the tidal distributary furnishes clear evidence of notable recent recession. Russell's description in 1891 indicates that the glacier was much farther out, and the general absence of vegetation on the margin of the fiord, as well as the freshly-smoothed